Unverified50% confidenceFactExact time

嵌入模型可将文本压缩为1536维或1024维的浮点向量，OpenAI的text-embedding-ada-002和Cohere的embed-multilingual是常见示例

1

Sources

50%

Confidence

Long-term

Relevance

7/17/2026

First Seen

Sources

编程Agent记忆缺失？SSH同步开源方案让AI真正记住你

Related Claims

VerifiedOpenAI 的 text-embedding-3-small 模型会将文本转换为 1536 维的浮点数数组81% similar Unverified现代Embedding模型（如OpenAI的text-embedding-ada-002、阿里的千问Embedding系列）能够将整个句子或段落编码为固定维度的稠密向量，通常为768维或1024维80% similar UnverifiedOpenAI的text-embedding-ada-002能将文字映射到高维向量空间，语义相近的文本在向量空间中距离更近80% similar

Cite This Claim

Stable URI

https://kongchang.com/claim/542454

API

curl https://kongchang.com/api/v1/knowledge/claims/542454

MCP

get_claim(id=542454)